Association for Public Policy Analysis & Management

This paper uses marijuana decriminalization laws, passed in 21 states over the last 40 years, to analyze the differences in earnings and employment that result from being arrested. Reduced form and, when the data allow, instrumental variable models are used to exploit the state-by-year variation in arrests that have resulted from states reducing penalities for marijuana possession. Data from the FBI’s Uniform Crime Reporting statistics and the Current Population Survey allow for age, gender and race specific estimates, which is critical considering the heterogeneity in rates of arrests across these delineations. Labor market outcomes in the CPS allow for an analysis of whether decriminalization laws affect extensive and intensive margins. Decreased penalties for marijuana possession are positively correlated with the probability of employment, although the results are imprecise. Additionally, there are non-trivial increases in weekly earnings for individuals living in states with decreased penalties, with the largest effects for black adults. This result is consistent with existing literature that suggests black adults, especially men, have the highest probability of being arrested for marijuana posssesion and therefore stand to benefit the most from removing these penalties.